Nagios Support Forum

Hi Guys,

I want to establish nagios distributed architecture.
Will you please help me?? I read articles on nsca. But I didn't got that.
Will you guys please explain it in simple manner.??


Thanks in advance..

NSCA allows Nagios to receive Passive check results from remote hosts.

The remote hosts must be using some sort of remote agent (like NSClient++ for Windows servers) to send the check results passively to Nagios.

I have a how-to guide here which may help:

http://sites.box293.com/nagios/guides/nsca

Does this help at all?

Hi,
Thanks for the quick reply. That link tells only how to create NSCA server.
I want to create distributed Nagios arcitecture, That means there will be one nagios server and multiple nagios as client.

Please guide me.

And thanks again that link is very helpful to understand NSCA.

Do you want multiple Nagios servers responding back to one central Nagios server using NSCA?

Is there any reason why you're wanting to use NSCA over other methods like Mod Gearman?

Hi,

Thanks for the reply. You got exactly right thing.
I want multiple Nagios servers responding back to one central Nagios server.
Here is my scenario.

I have already installed three nagios in my lab environment.
I want to make one nagios as central server and other will send there data to central nagios.
Other two are assumed to be established in different premises.

I don't want to to load balance the checks.

It would be great if central nagios also perform checks.

I tried mntos but my requirement is to access all host as they are local in nagios.
I do not need to login manually to other nagios.

will Mod Gearman fullfill this requirement.?? Or is there any other facility that will do exactly same with nagios core??

Please help.

Thanks in advance.

Both NSCA and Mod Gearman will meet these requirement but they both have different pros and cons.

NSCA will require defining host/service definitions at all locations. At your central location these will just be passive check objects. There is a lot more administrative overhead however it does work.

Mod Gearman will allow you to configure it all from a central site. You will only have one central nagios server and it will push out the checks to the "workers". The workers will be located at each separate site. They have all the required Nagios plugins installed to execute the checks locally and then send the results back to the central Nagios server. The workers can be configured to only process checks that are members of specific nagios host/service groups. So on your central Nagios server, your host and service definitions would have the hostgroup and servicegroup defined as well (like remote-site-01).

Have a look at this guide, it will get you up and running with a central Mod Gearman Job server and two additional worker servers. You will then need to configure your workers to target their respective Nagios groups (/etc/mod_gearman/mod_gearman_worker.conf).

http://sites.box293.com/nagios/guides/mod-gearman

Also, here is some more documentation on Mod Gearman:

https://labs.consol.de/nagios/mod-gearman/

It doesn't display in UI. does it?

It doesn't display in UI. does it?

Could you please clarify this question a bit? What were you asking about specifically.

Hi, I am able see worker nod in "gearman_top" view. But I don't understand hoe to distribute checks to worker node.
And also where to check received check result from worker node.

Please help.

Nagios is automatically using the worker node(s) to perform the checks, instead of Nagios doing the work.

A worker can be on the same server as the Nagios server, Nagios is just handing off the task to Mod Gearman.

When Nagios starts, if there is a line in nagios.cfg like this:
broker_module=/usr/lib64/mod_gearman/mod_gearman.o config=/etc/mod_gearman/mod_gearman_neb.conf
Then this is how Nagios is handing off the checks to Mod Gearman.

The workers check into to the Mod Gearman server and it hands the tasks to the workers.

And like I explained in a previous host, you can configure your workers to only perform checks for host/service objects if these objects are members of particular Nagios host/service groups.

So this way they workers can say "I'm only going to do checks for group_abc" and the Mod Gearman server will hand it only those checks.

Make sense?